1. Introduction 1. 1. R&D History of HDS Catalyst Natural petroleum resources, first exploited in the 19th century, have provided the cheap energy resources necessary for the recent unprecedented development in human society. It is no exaggeration to state that daily life depends completely on petroleum products. Petroleum is a mixture of hydrocarbons with many structures and molecular weights. In addition, petro- leum contains many molecules incorporating hetero- atoms such as sulfur, nitrogen and vanadium. The most common heteroatom-containing molecules are thiophene type compounds. Such sulfur-containing materials are indispensable in the fields of synthetic gum, agriculture and drugs, but also cause the most problems in the oil refining industry. In particular, the required level of desulfurization of petroleum products has been increasing gradually. Many methods have been developed for desulfurizing petroleum such as doctor treatment, copper sweetening and others. However, hydrodesulfurization (HDS) is the most effec- tive and feasible method, because sulfur is removed as H2S just by contacting sulfur-containing compounds with catalyst under a hydrogen atmosphere. Hydro- desulfurization is the ultimate chaotic chemical reac- tion, as it is to obtain mixture (product) from mixture (feed) with using mixture (catalyst). Nevertheless, higher levels of hydrodesulfurization to above 99.95 % are now required for the diesel fraction. Many types of HDS catalyst are now utilized in the oil industry depending on the feed properties and reac- tion conditions 1) . HDS catalysts are based on CoMo/ Al2O3 or NiMo/Al2O3 with various modifications, such as use of tungsten instead of molybdenum 2)～9) , additives such as silica 10)～15) , phosphorus 16)～24) , boria 13),22),25)～28) , fluorine 29)～35) , zeolite related materi- als 36)～42) , titania 13),43)～51) and others, and optimized preparation method and physicochemical properties such as shape, pore size distribution, and others 15),21),52)～61) . Many excellent reviews and texts on HDS have been published since the early 1950’s 13),20),62)～77) . Strangely enough, the origin of HDS catalysts is rarely described. One reason might be that HDS catalyst technologies have generally been developed by industrial concerns and the history of development may depend on com- mercial secrets. However, the origin of HDS catalysts is interesting for all HDS catalyst researchers. The origin lies in experimental investigations at the beginning of the 20th century 78),79) . In particular, studies on the hydrogenation of coal and heavy oil under high pres- sure by Friedrich Bergius and Carl Bosch in Germany during the 1910’s are important 80)～83) . These investi- 109 Journal of the Japan Petroleum Institute, 56, (3), 109-121 (2013) J. Jpn. Petrol. Inst., Vol. 56, No. 3, 2013 ＊ E-mail: [email protected][Review Paper] Regeneration of Residue Hydrodesulfurization Catalyst Ryuichiro IWAMOTO ＊ Technology and Engineering Center, Idemitsu Kosan Co., Ltd., 26 Anegasaki-kaigan, Ichihara, Chiba 299-0107, JAPAN (Received October 1, 2012) Reuse of residue hydrodesulfurization catalyst is very important, as a large volume of this catalyst is exchanged every year due to the high deactivation rate. However, stable regeneration of this catalyst is more difficult than for distillate HDS catalysts due to the more severe regeneration conditions and the presence of vanadium. Regeneration technologies of residue hydrodesulfurization catalyst are reviewed to assess the differences between regeneration of distillate HDS catalysts, mechanism of residue HDS catalyst regeneration, commercial regenera- tion, properties, structure, and activities of regenerated residue HDS catalyst, effect of vanadium and improvement of catalyst suitable for regeneration. The presence of vanadium accelerates both insufficient recovery of activity and reduced catalyst strength due to formation of NiMoO4 and Al2(SO4)3, respectively. However, removal of vanadium might be not essential. Vanadium acts indirectly as oxidative catalyst and increases the regeneration temperature and transformation of SOx into H2SO4. Therefore, milder regenerated conditions and control of the properties of used catalyst as well as improvement of catalyst may allow improved regeneration. Keywords Residue, Regeneration, Hydrodesulfurization catalyst, Vanadium, Nickel molybdenum oxide, Aluminum sulfate
Transcript
1. Introduction
1. 1. R&D History of HDS CatalystNatural petroleum resources, first exploited in the
19th century, have provided the cheap energy resources necessary for the recent unprecedented development in human society. It is no exaggeration to state that daily life depends completely on petroleum products. Petroleum is a mixture of hydrocarbons with many structures and molecular weights. In addition, petro-leum contains many molecules incorporating hetero-atoms such as sulfur, nitrogen and vanadium. The most common heteroatom-containing molecules are thiophene type compounds. Such sulfur-containing materials are indispensable in the fields of synthetic gum, agriculture and drugs, but also cause the most problems in the oil refining industry. In particular, the required level of desulfurization of petroleum products has been increasing gradually. Many methods have been developed for desulfurizing petroleum such as doctor treatment, copper sweetening and others. However, hydrodesulfurization (HDS) is the most effec-tive and feasible method, because sulfur is removed as H2S just by contacting sulfur-containing compounds with catalyst under a hydrogen atmosphere. Hydro-desulfuri zation is the ultimate chaotic chemical reac-
tion, as it is to obtain mixture (product) from mixture (feed) with using mixture (catalyst). Nevertheless, higher levels of hydrodesulfurization to above 99.95 % are now required for the diesel fraction.
Many types of HDS catalyst are now utilized in the oil industry depending on the feed properties and reac-tion conditions1). HDS catalysts are based on CoMo/Al2O3 or NiMo/Al2O3 with various modifications, such as use of tungsten instead of molybdenum2)～9), addit ives such as s i l ica10)～15), phosphorus16)～24), boria13),22),25)～28), fluorine29)～35), zeolite related materi-als36)～42), titania13),43)～51) and others, and optimized preparation method and physicochemical properties s u c h a s s h a p e , p o r e s i z e d i s t r i b u t i o n , a n d others15),21),52)～61).
Many excellent reviews and texts on HDS have been published since the early 1950’s13),20),62)～77). Strangely enough, the origin of HDS catalysts is rarely described. One reason might be that HDS catalyst technologies have generally been developed by industrial concerns and the history of development may depend on com-mercial secrets. However, the origin of HDS catalysts is interesting for all HDS catalyst researchers. The origin lies in experimental investigations at the beginning of the 20th century78),79). In particular, studies on the hydrogenation of coal and heavy oil under high pres-sure by Friedrich Bergius and Carl Bosch in Germany during the 1910’s are important80)～83). These investi-
109Journal of the Japan Petroleum Institute, 56, (3), 109-121 (2013)
Regeneration of Residue Hydrodesulfurization Catalyst
Ryuichiro IWAMOTO＊
Technology and Engineering Center, Idemitsu Kosan Co., Ltd., 26 Anegasaki-kaigan, Ichihara, Chiba 299-0107, JAPAN
(Received October 1, 2012)
Reuse of residue hydrodesulfurization catalyst is very important, as a large volume of this catalyst is exchanged every year due to the high deactivation rate. However, stable regeneration of this catalyst is more difficult than for distillate HDS catalysts due to the more severe regeneration conditions and the presence of vanadium. Regeneration technologies of residue hydrodesulfurization catalyst are reviewed to assess the differences between regeneration of distillate HDS catalysts, mechanism of residue HDS catalyst regeneration, commercial regenera-tion, properties, structure, and activities of regenerated residue HDS catalyst, effect of vanadium and improvement of catalyst suitable for regeneration. The presence of vanadium accelerates both insufficient recovery of activity and reduced catalyst strength due to formation of NiMoO4 and Al2(SO4)3, respectively. However, removal of vanadium might be not essential. Vanadium acts indirectly as oxidative catalyst and increases the regeneration temperature and transformation of SOx into H2SO4. Therefore, milder regenerated conditions and control of the properties of used catalyst as well as improvement of catalyst may allow improved regeneration.
gations revealed that some metal oxides or sulfides can hydrogenate sulfur-containing compounds and remove sulfur as a form of H2S under high-pressure hydrogen. This catalyst technology was adopted and further devel-oped by I. G. Farben, a German chemical company, and results were filed as patents after 192184)～89). For example, bulk molybdenum sulfide and cobalt oxide were reported to be effective catalysts. Notably, the synergic effect between cobalt and molybdenum was already described in these patents. Later, this catalyst technology was adopted by major American oil compa-nies such as Standard Oil New Jersey (now Exxon Mobil) and was finally established as the familiar con-ventional HDS catalyst.
Scientific investigations on HDS catalyst continued in the early 1930’s. HDS of shell oil over bulk CoS and MoS2 was reported in 193390). Thiophene HDS over bulk MoS2, CoS and Co3O4 was also reported in 193391). Steady sulfide state is very important to in-crease the number of active sites. However, the pre-sulfiding procedure had not become common at that time. Many materials were studied as support to ob-tain higher dispersion of molybdenum such as pumice, charcoal, clay minerals, brick, diatomaceous earth, bauxite, fuller’s earth, SiO2, MgO, Al2O3 and others62). Finally, thiophene HDS over the prototype of conven-tional Co_Mo/Al2O3 and NiMo/Al2O3 was described in 194392). The continued efforts of many researchers in industrial and academic fields have increased the activi-ty level of HDS catalyst by more than ten times in the past five decades93).
HDS was initially intended to reduce process corro-sion, eliminate the sour odor called skunk juice, or to improve the efficiency of tetra-alkyl lead used as an octane number booster for the gasoline fraction. However, a new use of HDS was recognized in the 1960’s, as public awareness of serious human health problems due to sulfur in petroleum was increased. The rapidly increasing consumption of petroleum based on postwar rehabilitation and high economic growth in Japan significantly increased the number of patients with bronchial asthma and chronic bronchitis around industrial areas. Research revealed that SOx from petroleum was the most important cause. Therefore, Japanese refineries introduced residue HDS units or vacuum gas oil HDS units to satisfy the regulations which restricted sulfur content in most fuel oils to below 0.5 %. The beneficial effect of sulfur reduction in fuel oil is quite clear, because the number of patients drastically decreased with the reduced SOx concentra-t i on in t he a tmosphe re a s shown in Fig. 194). Consequently, modern HDS catalysts have allowed more effective crude oil processing and higher economic growth without higher atmospheric concentration of SOx in Japan.
1. 2. Need for Regeneration of Residue HDS Catalyst
Human activity generates large amounts of various types of waste resulting in increasing and cumulative effects on the global environment. Therefore, many enterprises are trying to achieve zero emissions of waste through modification of manufacturing processes. The petroleum industries are also making great efforts to re-duce emissions of wastes and to build environmentally friendly refineries. HDS catalyst is essential to sup-port our modern petroleum-based society, and residue HDS catalyst is one of the most important types of waste from refineries, as large volumes of this catalyst are exchanged every year due to the high deactivation rate. Therefore, regeneration and reuse of residue HDS catalyst would significantly reduce waste from re-fineries. However, residue HDS catalyst is rarely re-generated, whereas HDS catalysts for distillate fractions such as light gas oil and kerosene are commonly regen-erated. Residue HDS catalyst is very difficult to sta-bly regenerate due to insufficient recovery of activity and decreased catalyst strength95)～102). Insufficient re-covery of catalytic activity shortens the catalyst life and decreased catalyst strength results in catalyst breakage during reloading into the reactor.
This review summarizes the technology of residue HDS regeneration and compares regeneration of distil-late HDS catalyst and residue HDS catalyst.
2. Basic Chemistry of Catalyst Regeneration
2. 1. Category of HDS RegenerationRegeneration of HDS catalysts is classified into on-
site regeneration and off-site regeneration (Fig. 2). In on-site regeneration, the catalyst is regenerated inside the reactor without removal using existing process equipment. In off-site regeneration, the catalyst is re-generated outside the reactor after removal using spe-
110
J. Jpn. Petrol. Inst., Vol. 56, No. 3, 2013
Fig. 1● Relationship between SOx Level and Number of Patients with Bronchial Asthma and Chronic Bronchitis in Yokkaichi, Japan (from White Paper of the Environment Div., Yokkaichi City94))
cial regeneration equipment. Recently, off-site regen-eration has become common because of several advantages for refineries as follows:
-better activity recovery with precise temperature control,
-reduced fines and scales which cause maldistribu-tion and pressure drop during the hydrotreating reac-tion,
-separation of unregenerable catalyst as well as unsuccessfully regenerated catalyst,
-reduced risk of reactor damage due to unsuitable heat control,
-shorter shut down maintenance period.Furthermore, off-site regeneration can be classified
in to oxidat ive regenera t ion and re juvenat ion . Oxidative regeneration is intended to remove only coke from the active sites by combustion. In some cases, postoxidative regeneration treatment is incorporated to obtain further recovery of active metal dispersion with various organic agents such as citric acid, tartaric acid, oxalic acid, malonic acid, butanediolglycolic aldehyde, acetaldol, various glycols, etc.103). On the other hand, rejuvenation is intended to achieve complete regenera-tion into the original structure by removing both coke and contaminant metals such as vanadium and nickel with chemical reagents. Therefore, the main target of rejuvenation is residue HDS catalysts. However, extraction of only contaminant metals but selectively retaining act ive molybdenum is very difficul t . Therefore, rejuvenation is still at the stage of research and development. This review mainly discusses oxi-dative regeneration.
2. 2. Differences between Residue and Distillate HDS Catalyst Regeneration
Table 1 shows the relationships between target feed, reaction conditions, and resultant used catalyst proper-ties. The residue HDS process must remove not only sulfur but also contaminant metals such as vanadium and nickel from the feedstock, so requires several spe-cific catalysts such as hydrodemetallization (HDM), HDS and intermediate state catalysts. On the other hand, the distillate HDS process normally requires only a single HDS catalyst. Residue HDS requires severer reaction conditions to achieve the same HDS rate as distillate HDS catalysts, so that lower LHSV and higher catalyst loading result in shorter operation period which means higher catalyst cost. HDS of distillate fractions such as naphtha and light gas oil need milder reaction conditions, so used catalysts contain low amounts of coke. In contrast, used residue HDS catalyst contain relatively higher amounts of coke. Furthermore, the coke burning temperature of residue HDS catalyst is also higher than that of distillate HDS catalysts. In addition, residue HDS catalysts accumulate vanadium and nickel which are not observed in distillate HDS cat-alysts. These differences lead to technical problems such as insufficient recovery of activity and decrease of catalyst strength. In fact, less residue HDS catalyst is regenerated and most used catalyst goes to metal recov-ery, whereas more than half of distillate HDS catalysts are commonly regenerated as shown in Fig. 3104).2. 3. Chemistry of Residue HDS Catalyst102)
Deactivation mechanisms occurring during the hydrotreating reaction have been extensively studied by
111
J. Jpn. Petrol. Inst., Vol. 56, No. 3, 2013
Fig. 2 Classification of HDS Catalyst Regeneration
Table 1 Relationship between Feed and Properties of Regenerated Catalysts
CatalystFeed properties Reaction condition Properties of used catalyst
many researchers110)～113), but deactivation mechanisms during regeneration are less well known, especially for residue HDS catalyst. Reuse of residue HDS catalyst requires precise understanding of the regeneration be-havior. The oxidative regeneration mechanism of hy-drotreating catalyst was previously reviewed by Frimsky109),110). Coke and contaminant metals (Ni, V and Fe) accumulate on residue catalyst during the hy-drotreating reaction. Normally, such metals are pres-ent as sulfides, as are active metals such as MoS2 and Co9S8. For example, vanadium is considered to accu-mulate as V3S4
111). These metal sulfides and coke are oxidized by the regeneration process.
The oxidative regeneration reaction consists of the following equations95),112).
C＋O2＝CO2 (1)
2C＋O2＝2CO (2)
4Horg＋O2＝2H2O (3)
MSx＋yO2＝MOZ＋xSO2 (4)
where Horg refers to the hydrogen contained in the coke, and M refers to metals on the catalyst such as Co, Ni, Mo, W, V and Fe. During oxidative regeneration, the states of the coke, active metals, accumulated metals, and support change significantly96)～99),112). Ac t ive me ta l spec ie s fo rm β -NiMoO4, MoO3, Al2(MoO4)3, FeMoO4 and NiAl2O4 depending on the regeneration condition and nature of the catalyst. Alumina support may also transform from γ-Al2O3 to δ-Al2O3 or α-Al2O3 due to the exothermal reaction of coke combustion98). Accumulated vanadium sulfide on catalyst also transforms to V2O5 by oxidation98).
Regeneration temperatures which are too high will not completely restore the activity due to sintering of the active metals, even if the specific surface area is re-stored to that of fresh catalyst95). Vanadium accumula-tion may prevent restoration of activity due to blocking of the pore structure and active sites113), and vanadium
may also prevent redispersion of the active metals96). On the other hand, regeneration under certain condi-tions leads to even higher HDS activity compared to fresh catalyst because of the greater dispersion of the active metal114),115). Therefore, the chemistry of HDS catalyst regeneration is not yet well understood. In particular, the effect of vanadium accumulation on the regeneration of residue HDS catalyst has not been clari-fied.
3. Commercial Regeneration Process
Several companies worldwide such as EUROCAT, TRICAT, POROCEL, Nippon Ketjen and RASA (previ-ous NCRI) provide commercial regeneration services with proprietary regeneration processes intended for re-generating distillate HDS catalysts and residue HDS catalysts. For example, Fig. 4 shows the schematic commercial regeneration process of RASA which adopts a specific belt conveyor system. The catalyst is loaded on the slowly moving belt as a shallow bed. The regeneration process mainly consists of a stripping step and main regeneration step. In the first step, pro-cess oil absorbed on the catalyst is stripped by heating under the no oxygen condition. This step is very important to prevent undesirable exothermal reaction during the following main regeneration step. Then, catalyst without process oil is carefully regenerated under suitably controlled temperature and oxygen con-ditions. During the operation, catalyst properties are checked occasionally and the results are fed back to adjust the optimum regeneration conditions. The final regenerated catalyst is screened into desired sizes and packed in a flexible container bag.
112
J. Jpn. Petrol. Inst., Vol. 56, No. 3, 2013
Fig. 3● Regeneration of HDS Catalysts in Japan (from “Hydro-processing ̶Science & Technology,” ed. by Kabe, T., IPC, (2000), p. 407104).)
Fig. 4● Commercial Regeneration Process of RASA (presented by RASA Industries Ltd.)
4. Structural Description of Regenerated Residue HDS Catalyst102)
Typical properties of the catalysts for regeneration depend on the histories during commercial operation as shown in Table 2. The catalysts are obtained from different parts of the reactor. In general, catalyst loaded at the upper part of reactor shows higher accumulation of vanadium and nickel. Such catalysts can be regen-erated in the laboratory under steady temperature con-ditions from 573 K in dry air to eliminate deposited carbon and sulfur. The total amount of vanadium and nickel accumulation based on fresh catalyst during the hydrotreating reaction is defined as Metal on Catalyst (MOC). Catalyst with higher MOC tends to show higher sulfur content and lower surface area116)～118). This is not due to incomplete regeneration, because the coke on the catalyst was reduced to an adequately low level.
Figure 5 shows the effect of MOC on the structure of the regenerated catalyst measured by X-ray diffraction (XRD)102). The fresh catalyst showed only broad peaks belonging to the γ-alumina support, and no peaks for nickel or molybdenum compounds. However, the regenerated catalysts with different MOC showed peaks assigned to NiMoO4, which indicates the presence of aggregated nickel and molybdenum components. The peak intensities of NiMoO4 increased with higher
MOC. Aggregation of active metal species may occur during the hydrotreatment reaction and/or during the re-generation procedure119). However, the regeneration procedure might be the main contributor to the aggrega-tion of active metals, since aggregation of Ni and Mo sulfides was not obvious in the used catalyst as shown in Fig. 8.
Other peaks for Al2(SO4)3 were observed above 2.5 wt% of MOC and increased with higher MOC. This observation well agrees with the finding that sulfur on the regenerated catalyst increased with higher MOC. Since catalyst strength decreases with MOC, this loss of strength may be related to the transformation of alu-mina into Al2(SO4)3 during the regeneration procedure.
Figure 6 shows the electron probe micro analysis (EPMA) line analysis of aluminum, molybdenum, nickel, vanadium and sulfur for the regenerated catalyst. Nickel, molybdenum and aluminum had almost uni-form distributions over the regenerated catalyst as well as the original fresh catalyst.
In contrast, the distributions of sulfur (Al2(SO4)3) and vanadium were quite similar and were located around the exterior of the catalyst particles. The sulfur might originate from metal sulfides such as MoS2 and V3S4. However, if the sulfur in metal sulfides reacts with alu-mina directly, Al2(SO4)3 could form anywhere over the catalyst as metal sulfides are present all over the cata-lyst particles. One possibility is that SOx from metal
a) Vanadium and nickel accumulation calculated based on fresh catalyst.b) Precise quantitative analysis is ＞1.0 wt%.
Fig. 5● Effect of Metal on Catalyst (MOC) on the Structure of Regenerated Catalyst
Fig. 6 EPMA Line Analysis of the Regenerated Catalyst
sulfides is catalytically transformed into H2SO4 by vanadium, which is well known to be an active oxida-tion catalyst. Then, the H2SO4 subsequently reacts with alumina so that the location of sulfur is associated with the location of vanadium. The reaction can be described by Eqs. (3), (4) and the following equations.
SO2＋1/2O2＝SO3 (5)
SO3＋H2O＝H2SO4 (6)
3H2SO4＋Al2O3＝Al2(SO4)3＋3H2O (7)
This type of Al2(SO4)3 formation might be unique to the regeneration of residue HDS catalyst, as such a phe-nomenon has rarely been observed in distillate HDS catalyst regeneration.
5. Ef fec t o f Regenerat ion Temperature on Properties of Residue HDS Catalyst120)
Figure 7 shows the carbon, sulfur and nitrogen con-tents of HDS catalysts regenerated at different tempera-tures. The compositions show drastic changes depending on the calcined temperature. These compo-nents started to decrease above 573 K and almost disap-peared at 673 K.
Figure 8 shows the XRD patterns of the residue HDS catalysts regenerated at different temperatures, as well as the fresh and used catalysts. The fresh catalyst and used catalysts mainly showed only few broad peaks assigned to the γ-alumina support. No evidence of NiMoO4 formation, indicating the aggregation of nickel and molybdenum components, was observed in these catalysts. The regenerated catalysts showed no dis-tinct differences up to 623 K. However, the catalyst regenerated at 723 K showed clear formation of NiMoO4. Therefore, the aggregation of Ni and Mo should be related to higher temperature121). To con-firm the above discussion more theoretically, the effect of coke burning on catalyst temperature was calculated. Oxidative regeneration can be described by the follow-ing equations.
C＋1/2O2→CO＋394 kJ (7)
H2＋1/2O2→H2Og＋243 kJ (8)
If the H/C molar ratio of coke on the catalyst is 0.8 and the specific heat of alumina-based catalyst is 1.1 kJ/kg K, the 1 wt% of coke burning in adiabatic reaction will increase the temperature of the catalyst by about 40 K. Therefore, rapid coke burning and heat release restriction could accelerate the aggregation of nickel and molybdenum. Figure 9 shows the thiophene HDS activity for catalysts regenerated at different tem-peratures. The activity increased with higher regener-ated temperature up to 673 K due to coke elimination, and then started to decrease above 723 K. This tem-perature corresponds well to the aggregation of active
metals as shown in Fig. 8.
6. A c t i v i t y o f R e g e n e r a t e d R e s i d u e H D S Catalyst102)
Figure 10 shows the relationship between MOC and HDS activity for the fresh and regenerated residue HDS catalysts. The HDS activity gradually decreased with MOC. Based on the Arrhenius plot, regenerated cata-lysts have almost the same activation energy as fresh catalyst (ca. 25,000 cal/mol), with no significant change associated with higher MOC.
Figure 11 shows the relationship between MOC and hydrodenitrogenation (HDN) activity. HDN activity also decreased with higher MOC. Similarly, regener-ated catalysts had almost same activation energy as fresh catalyst (ca. 18,000 cal/mol), with no significant change associated with higher MOC.
Figures 12 and 13 show the relationships between MOC and hydrodevanadiumat ion (HDV), and
114
J. Jpn. Petrol. Inst., Vol. 56, No. 3, 2013
Fig. 7● Effect of Regeneration Temperature on Carbon, Sulfur and Nitrogen Content on the Catalyst
hydrodenickelation (HDNi) activities, respectively. In contrast to the HDS and HDN findings, the HDV and HDNi activities may increase with higher MOC (activa-tion energy: ca. 13,000 cal/mol). This result suggests that the accumulated vanadium sulfide forms new active HDM sites, and may enhance the scission of the V_C bond in heavier molecular fractions and trap more vanadium from the feed. Vanadium accumulation on regenerated catalyst has been also reported to increase the selectivity of HDM122).
Figure 14 similarly shows that the hydrodeasphaltene (HDAs) activity increased with higher MOC, similarly to the HDV and HDNi activities (activation energy: ca. 17,000 cal/mol).
Since the formation of aggregated NiMoO4 increases with higher MOC, deactivation of HDS, HDN and hydrodemicrocarbon residue (HDMCR) should be mainly attributed to the decreased number of active
115
J. Jpn. Petrol. Inst., Vol. 56, No. 3, 2013
Fig. 8● X-ray Diffraction Patterns of Fresh, Used and Regenerated Catalysts (MOC＝1.8 %)
Fig. 9● Thiophene HDS Activity of Catalysts Regenerated at Different Temperatures (MOC＝1.8 %)
Fig. 10 HDS Activity of Fresh and Regenerated Catalysts
Fig. 11 HDN Activity of Fresh and Regenerated Catalysts
Fig. 12 HDV Activity of Fresh and Regenerated Catalysts
Fig. 13 HDNi Activity of Fresh and Regenerated Catalysts
sites. The HDS reaction pathway may not change with higher MOC as the activation energy does not obviously change between the fresh and regenerated catalysts as described above. The effect of nickel accumulation is rather difficult to estimate, but might act as a poison for the HDS reaction, as the nickel content of fresh HDS catalyst is already optimized and excess nickel loading on the catalyst tends to decrease HDS activity.
7. Effect of Vanadium on Residue HDS Catalyst
As described above, vanadium affects the active metal dispersion and catalyst strength of residue HDS cata-lyst. To quantitatively investigate the effect of vanadium on NiMoO4 aggregation, Micro-XRD patterns through-out the section of catalyst particles were measured (Catalyst with MOC: 2.8 wt% and regenerated temper-ature: 723 K was used.). Micro-XRD can provide
precise information on the structural change through a specific catalyst particle from outside to inside, whereas conventional XRD measurement provides only the average structure of the catalyst123). The amount of NiMoO4 over the catalyst particle was plotted using the intensity ratio of NiMoO4 peak/Al2O3 peak in Fig. 15. The findings clearly indicated that the formation of NiMoO4 linearly increased towards the center of the catalyst particle. Vanadium was mainly located on the outside of the catalyst, so the amount of NiMoO4 for-mation was not correlated with the location of vanadium. This result seems not agree with the above findings that the formation of NiMoO4 is correlated with the vanadium content. However, vanadium may affect NiMoO4 for-mation indirectly in the following ways.(1) Vanadium may not directly affect the aggregation of nickel and molybdenum through certain chemical inter-action.
In fact, formation of NiMoO4 is also observed on re-generated distillate HDS catalyst which contains no vanadium115). This observation may indicate that vanadium is indirectly involved in NiMoO4 formation.(2) Vanadium acts as an oxidation catalyst to accelerate coke burning, resulting in increased temperature inside the catalyst particle. (3) Higher vanadium accumulation on the catalyst has less refractory coke, as this catalyst is located at the upper part of reactor and exposed to lower reaction tem-peratures112),124)～129). Therefore, coke burning is easier and leads to increased temperature inside the catalyst particle.(4) Pore mouth plugging by vanadium prevents heat re-lease from inside the catalyst.
Figure 16 shows the schematic mechanism of resi-
116
J. Jpn. Petrol. Inst., Vol. 56, No. 3, 2013
Fig. 14 HDAs Activity of Fresh and Regenerated Catalysts
(a) NiMoO4/Al2O3 intensity ratio calculated from Micro-XRD, (b) vanadium content.
Fig. 15 Distribution of NiMoO4 and Vanadium Content through the Catalyst Particle
due HDS regeneration. Vanadium accumulation accelerates the formation of NiMoO4 and leads to reduced catalytic activities except for HDM and HDAs. Vanadium accumulation also accelerates the transfor-mation of alumina support into Al2(SO4)3 and decreases the strength of catalyst, especially based on abrasion.
Coke accumulation might also affect the regeneration behavior and resulting catalytic activity. Figure 17 shows the schematic image of expected regeneration conditions associated with regeneration temperature. If the temperature is too low, coke cannot be eliminated. However, if the temperature is too high, active site aggregation occurs. Therefore, careful control of coke combustion to prevent increases in catalyst temperature is very important to achieve maximum recovery of activity. This precaution can also help to prevent the formation of Al2(SO4)3 which leads to reduced catalyst strength. In summary, residue HDS catalysts can be regenerated even in the presence of vanadium by con-trolling the regeneration conditions and properties of the used catalyst. Removal of vanadium is not essen-tial to regenerate residue HDS catalyst.
8. Estimation of Lifetime for Regenerated Residue HDS Catalyst130)
Estimation of the lifetime of regenerated residue HDS catalyst is very important to achieve steady opera-tion in commercial processes. If the exact lifetime of regenerated catalyst is not clear, operators cannot have
confidence in using the regenerated catalysts . Figure 18 shows the relationship between MOC and HDS activity. HDS catalyst deactivates in three steps105)～107). At the initial stage of reaction, the cata-lyst is deactivated due to coke deposition inside the catalyst pore, and in the middle stage of reaction, the catalyst is deactivated slowly due to metal deposition and hard coke transformation. In the final stage, the catalyst is deactivated rapidly again due to pore mouth plugging by metal deposition. The limited MOC of each catalyst at which pore mouth plugging occurs is defined as the Metal Capacity (MC). Regenerated catalyst lifetime might also depend on the MC. To explain in more detail, the state of the catalyst during the hydrotreating reaction using fresh and regenerated catalyst is summarized in Fig. 19. During first com-mercial operation using fresh catalyst, the activity decreased gradually by coke and metal deposition to the level of minimum activity required for commercial op-eration (Fig. 20(a)). Regeneration then restores the activity to a certain level by eliminating the coke in the catalyst pores, but the metal plugging the pore mouth remains (Fig. 20(b)). During second commercial operation using the regenerated catalyst, the activity is decreased again by coke and further metal deposition (Fig. 20(c)). In principle, only the catalyst without fully plugged pore mouths can be used as regenerated catalyst. If total metal accumulation on the fresh cata-lyst (actual) and regenerated catalyst (prediction) is below the MC (Fig. 19(a)), the regenerated catalyst can be used for commercial operation as the activity will not decrease rapidly due to pore mouth plugging during the operation. However, if the total estimated MOC is
117
J. Jpn. Petrol. Inst., Vol. 56, No. 3, 2013
Fig. 16● Schematic Effect of Vanadium on Residue HDS Catalyst Regeneration
Fig. 17 Schematic Image of Reasonable Regeneration Conditions
Fig. 18 Relationship between Metal on Catalyst and HDS Activity
above the MC, such regenerated catalyst cannot be used for further commercial operation (Fig. 19(b)) . Therefore, it is very important to understand the MC, which can be estimated both from the pore plugging simulation and from long term pilot test.
9. Improvement of Residue HDS Catalyst Suitable for Regeneration131)～134)
Present regeneration technology cannot regenerate all residue HDS catalysts. MOC tends to be higher at the upper part of the reactor and gradually decreases in the lower part. Therefore, catalyst from the upper part is
118
J. Jpn. Petrol. Inst., Vol. 56, No. 3, 2013
(a) State after first commercial operation, (b) state after regeneration, (c) state after second commercial operation.
Fig. 19● Schematic Catalyst State during Hydrotreating Reaction Using Fresh and Regenerated Catalysts
(a) XRD patterns of regenerated catalysts, (b) relative catalyst strength based on reciprocal abrasion.
Fig. 20● Effect of Magnesium Addition on Catalyst Structure and Strength of Regenerated Catalyst
more difficult to regenerate. The main technological problems of regeneration for upper part catalyst depend on the catalyst strength based on abrasion rather than activity recovery. To increase the available amount of regenerated catalyst, improvements to prevent decreas-ing catalyst strength during regeneration were investi-gated. As described above, reduced catalyst strength is related to the formation of Al2(SO4)2 due to reaction between H2SO4 and the alumina support. Therefore, it is effective to trap H2SO4 before interaction with the alumina support. Alkali earth is an effective additive for this purpose. Chemical analysis and XRD mea-surement have proved that the addition of magnesium increases sulfur on catalyst, decreases the formation of H2SO4 and increases the strength of catalyst as shown in Fig. 20. This improved catalyst was loaded in a commercial residue HDS unit and used for one year. The regenerated catalyst could be reused in the com-mercial unit again without problems.
10. Conclusion
Regeneration of residue HDS catalyst is more diffi-cult than that of distillate HDS catalyst due to the accu-mulation of vanadium species on the catalyst during hydrotreating, which leads to the formation of NiMoO4 and Al2(SO4)3 during regeneration. NiMoO4 decreases the number of active sites and Al2(SO4)3 decreases cata-lyst strength. Therefore, HDS, HDN, and HDMCR activities and catalyst strength decrease gradually with higher MOC. On the other hand, the HDM and HDAs activities of regenerated residue HDS catalyst increased with greater metal accumulation. Vanadium may not chemically affect the active metal states, but the oxida-tion activity of vanadium accelerates coke burning and increasing temperature leads to aggregation of active metals. Milder regeneration conditions such as lower temperature and slower regeneration may prevent both reduced activities and lower catalyst strength. The addition of alkali earth metal to the catalyst in order to trap H2SO4 can prevent the formation of Al2(SO4)3 and help maintain catalyst strength. Such an effective technological counter measurement could help improve the regeneration of residue HDS catalysts.
AcknowledgmentThis study was partly supported under the auspices
of the Japan Petroleum Energy Center (JPEC). The author thanks Professor Atsushi Yamazaki, Waseda University, School of Creative Science and Engineering, Depar tment o f Resources and Envi ronmenta l Engineering with help in Micro-XRD measurements. The author also thanks RASA Industries Ltd. for pro-viding process flow data for this review.
References
1) Rhodes, A. K., Oil & Gas J., 2, 3 Oct., 5 (1995). 2) Maitra, A. M., Cant, N. W., Trim, D. L., Appl. Catal., 27, 9
(1986). 3) Breysse, M., Bachelier, J., Bonnelle, J. P., Cattenot, M.,
Cornet, D., Decamp, T., Duchet, J. C., Durand, R., Engelhard, P., Frety, R., Gachet, C., Geneste, P., Grimblot, J., Gueguen, C., Kasztelan, S., Lacroix, M., Lavalley, J. C., Leclercq, C., Moreau, C., De Mourgues, L., Olive, J. L., Payen, E., Portefaix, J. L., Toulhoat, H., Vrinat, M., Bull. Soc. Chim. Bel., 96, 829 (1987).
4) Kim, C.-H., Yoon, W. L., Lee, I. C., Woo, S. I., Appl. Catal. A: General, 144, 159 (1996).
5) Reinhoudt, H. R., van Langeveld, A. D., Kooyman, P. J., Stockmann, R. M., Prins, R., Zandbergen, H. W., Moulijn, J. A., J. Catal., 179, 443 (1998).
6) Ramírez, J., Gutiérrez-Alejandre, A., Catal. Today, 43, 123 (1998).
7) Vissenberg, M. J., van der Meer, Y., Hensen, E. J. M., de Beer, V. H. J., van der Kraan, A. M., van Santen, R. A., van Veen, J. A. R., J. Catal., 198, 151 (2001).
8) Reinhoudt, H. R., Troost, R., van Langeveld, A. D., van Veen, J. A. R., Sie, S. T., Moulijn, J. A., J. Catal., 203, 509 (2001).
9) Zuo, D., Vrinat, M., Nie, H., Maugé, F., Shi, Y., Lacroix, M., Li, D., Grange, P., Catal. Today, 93-95, 751 (2004).
10) Ahuja, S. P., Derrien, M. L., Le Page, J. F., Ind. Eng. Chem., Prod. Res. Devlop., 9, 272 (1970).
11) Massoth, F. E., Dhar, G. M., Shabtai, J., J. Catal., 85, 44 (1984). 12) Henker, M., Wendlandt, P. P., Valyon, J., Bornamann, P., Appl.
Catal., 69, 205 (1991). 13) Dhar, G. M., Srinivas, B. N., Rana, M. S., Kumar, M., Maity, S.
K., Catal. Today, 86, 45 (2003). 14) Kunisada, N., Choi, K.-H., Korai, Y., Mochida, I., Nakano, K.,
Appl. Catal. A: General, 273, 287 (2004). 15) Al-Dalama, K., Stanislaus, A., Energy & Feuls, 20, 1777 (2006). 16) Atanasova, D., Halachev, T., Uchytil, J., Kraus, M., Appl.
Catal., 38, 235 (1988). 17) Mangnus, P. J., van Veen, J. A. R., Eijsbouts, S., De Beer, V. H.
J., Moulijn, J. A., Appl. Catal., 61, 99 (1990). 18) Jian, M., Prins, R., Catal. Lett., 35, 193 (1995). 19) Kwak, C., Kim, M. Y., Choi, K., Moon, S. H., Appl. Catal. A:
General, 185, 19 (1999). 20) Iwamoto, R., Grimblot, J., Adv. Catal., 44, 417 (2000). 21) Sun, M., Nicosia, D., Prins, R., Catal. Today, 86, 173 (2003). 22) Ferdous, D., Dalai, A. K., Adjaye, J., Appl. Catal. A: General,
260, 153 (2004). 23) Maity, S. K., Ancheyta, J., Rana, M. S., Rayo, P., Catal. Today,
109, 42 (2005). 24) Iwamoto, R., Kagami, N., Sakoda, Y., Iino, A., J. Jpn. Petrol.
Inst., 48, (6), 351 (2005). 25) Tsai, M. C., Chen, Y. W., Kang, B. C., Wu, J. C., Leu, L. J.,
Ind. Eng. Chem. Res., 30, 1801 (1991). 26) Ramírez, J., Castillo, P., Cedeño, L., Cuevas, R., Castillo, M.,
Palacios, J., López-Agudo, A., Appl. Catal. A: General, 132, 317 (1995).
27) Dubois, J. L., Fujieda, S., Catal. Today, 29, 191 (1996). 28) Usman, A., Kubota, T., Araki, Y., Ishida, K., Okamoto, Y., J.
Catal., 227, 523 (2004). 29) Muralidhar, G., Massoth, F. E., Shabtai, J., J. Catal., 85, 44
(1984). 30) Jiratova, K., Kraus, M., Appl. Catal., 27, 21 (1986). 31) Papadopoulou, C., Lycourghiotis, A., Grange, P., Delmon, B.,
Appl. Catal., 38, 255 (1988). 32) Kwak, C., Lee, J. J., Bae, J. S., Choi, K., Moon, S. H., Appl.
Catal. A: General, 200, 233 (2000). 33) Song, C. J., Kwak, C., Moon, S. H., Catal. Today, 74, 193
119
J. Jpn. Petrol. Inst., Vol. 56, No. 3, 2013
(2002). 34) K i m , H . , L e e , J . J . , M o o n , S . H . , A p p l . C a t a l . B :
A: General, 301, 241 (2006). 36) Cid, R., Orellana, F., Agudo, A. L., Appl. Catal., 32, 327 (1987). 37) Welters, W. J. J., de Beer, V. H. J., van Santen, R. A., Appl.
Catal. A: General, 119, 253 (1994). 38) Fujikawa, T., Chiyoda, O., Tsukagoshi, M., Idei, K., Takehara,
Today, 43, 203 (1998). 58) Stanislaus, A., Al-Dolama, K., Absi-Halabi, M., J. Mol. Catal.
A, 181, 33 (2002). 59) Rana, M. S., Ancheyta, J., Rayo, P., Maity, S. K., Catal. Today,
98, 151 (2004). 60) Nicosia, D., Prins, R., J. Catal., 231, 259 (2005). 61) Iwamoto, R., Kagami, N., Iino, A., J. Jpn. Petrol. Inst., 48, (4),
237 (2005). 62) McKinley, J. B., “Catalysis Vol. 5, ” ed. by Emmett, P. H.,
Reinhold Publishing Co., New York (1957), p. 405. 63) Schuit, G. C., Gates, B. C., AIChE J., 19, 417 (1973). 64) Massoth, F. E., Adv. Catal., 27, 265 (1978). 65) Delmon, B., Bull. Soc. Chem. Belg., 88, 979 (1979). 66) Ratnasamy, P., Sivasanker, S., Catal. Rev.-Sci. Eng., 22, 401
(1980). 67) Furimsky, E., Catal. Rev.-Sci. Eng., 22, 371 (1980). 68) Vrinat, M., Appl. Catal., 6, 137 (1983). 69) Zdrazil, M., Catal. Today, 3, 387 (1988). 70) Prins, R., de Beer, V. H. J., Somorjai, G. A., Catal. Rev.-Sci.
Eng., 31, 1 (1989). 71) Topsøe, H., Clausen, B. S., Massoth, F. E., “Hydrotreating
Catalysis,” Springer, New York (1996). 72) Anderson, J. R., Boudart, M. (eds.), “Catalysis: science and
technology,” vol 11, Springer, New York (1996). 73) Grange, P., Vanhaeren, X., Catal. Today, 36, 375 (1997). 74) Eijsbouts, S., Appl. Catal. A: General, 158, 53 (1997). 75) Song, C., Catal. Today, 86, 211 (2003). 76) Bej, S. K., Maity, S. K., Turaga, U. T., Enegy & Fuels, 18, (5),
1227 (2004). 77) Mochida, I., Choi, K.-H., J. Jpn. Petrol. Inst., 47, (3), 145
(2004). 78) Sabatier, P., Senderenns J. B., Compt. Rend., 137, 1025 (1900). 79) Ipatieff, V. N., J. Russ. Phys. Chem. Soc., 38, 75 (1906). 80) Bergius, F., Angewandte Chemie, 34, 341 (1921). 81) Well, S. W., Energy & Fuels, 8, 415 (1994). 82) Bergius, F., “Chemical reactions under high pressure, Nobel
83) Bosh, C., “The Development of the Chemical High Pressure Method During the Establishment of the New Ammonia Industry,” (1932), http://nobelprize.org/nobel-prizes/chemistry/laureates/1931/bosch-lecture.html.
84) Bergius, F., GB Pat. 244730 (1926). 85) Bergius, F., US Pat. 1592772 (1926). 86) Bergius, F., Loeffler, S., GB Pat. 192849 (1923). 87) Bergius, F., Loffler, S., GB Pat. 192850 (1923). 88) Bergius, F., US Pat. 1344671 (1920). 89) IG Farbenindustrie AG, GB Pat. 315439 (1929). 90) Nametkin, S. S., Sanin, P. I., Makover, S. V., Tzuiba, A. N., J.
Appl. Chem. (USSR), 6, 494 (1933). 91) Pease, R. N., Keigton, W. B., J. Ind. Eng. Chem., 25, 1012
(1933). 92) Byrns, A. C., Bridley, W. E., Lee, M. W., Ind. Eng. Chem., 35,
121 (2004). 109) Furimsky, E., Massoth, F. E., Catal. Today, 17, 537 (1993). 110) Marafi, M., Stanislaus, A., Furimsky, E., “Handbook of spent
hydroprocessing catalysis, regeneration, rejuvenation and rec-lamation,” Elsevier, (2010).
120
J. Jpn. Petrol. Inst., Vol. 56, No. 3, 2013
111) Callejas, M. A., Martínez, M. T., Fierro, J. L. G., Rial, C., Jimenéz-Mateos, J. M., Gómez-García, F. J., Appl. Catal. A: General, 220, 93 (2001).
112) George, Z. M., Mahammed, P., Tower, R., Proc. 9th Int. Congr. Catal., (1988), p. 230.
113) Alvarez, D., Galiasso, R., Andréu, P., J. Jpn. Petrol. Inst., 22, (4), 234 (1979).
114) Arteaga, A., Fierro, J . L. G., Grange, P. , Demon, B. , Proceeding of symposium on advances in hydrotreating, ACS in Denver meeting, (1987), p. 339.
115) Oh, E. S., Yong-Chul Park, Y. C., Lee, I. C., Rhee, H. K., J. Catal., 172, 314 (1997).
116) Iwamoto, R., Sakota, H., Sugama, Y., 54th Annual Meeting of Jpn. Petrol. Inst., Tokyo, (2005), A11.
117) Iwamoto, R., Kagami, N., Eguchi, S., 55th Annual Meeting of Jpn. Petrol. Inst., Tokyo, (2006), A16.
118) Iwamoto, R., Kagami, N., Eguchi, S., 36th Petroleum-Petrochemical Symposium of Jpn. Petrol. Inst., Kagoshima, (2006), 1E17.
119) Podraza, F., Fuentes, S., Vrinat, M., Lacroix, M., Catal. Lett., 62, 121 (2007).
120) Iwamoto, R., Matsui, K., Yamazaki, A., J. Jpn. Petrol. Inst., 52, (4), 211 (2009).
121) Matsui, K., Yamazaki, A., Iwamoto, R., 37th Petroleum-Petrochemical Symposium of Jpn. Petrol. Inst., Sapporo, (2007), 1E17.
